Electronic Theses and Dissertations


Paige Potts



Document Type


Degree Name

Doctor of Philosophy



Committee Chair

Daniel Baker

Committee Member

Abby Parrill

Committee Member

Judith A Cole

Committee Member

Thomas R Sutter

Committee Member

Gary L Emmert


The diversity of the known and presumed physiological roles of the members of the GPCR superfamily motivate studies to expand both pharmacological and general knowledge. Computational GPCR ligand discovery/interaction studies require either a protein structure or a large number of known ligands. There are 120 unique GPCR with structures deposited in the Protein Data Bank (August 2021), leaving many GPCR without experimentally-determined structures. Additionally, many GPCR are orphan receptors lacking known endogenous ligands (potentially lacking other known synthetic small molecule ligands). Chapter I provides a brief overview of two common GPCR ligand discovery methods—homology modeling and pharmacophore modeling. Each of these topics are discussed in further detail in subsequent chapters. Chapter I also provides an overview of the various online databases that are commonly used for ligand identification studies and the differences between them. The benchmark study in chapter II examines the performance of a global versus local similarity measure applied to comparative modeling template selection for 6 previously crystallized, class A GCPR with the long-term goal of optimizing GPCR ligand identification efforts. Overall, the data supports the use of local similarity measures to guide template selection in protocols using comparative models to investigate ligand-receptor interactions. The in silico study in chapter III assesses the impact of key pharmacophore modeling user decisions that arise when known ligand numbers for a target of interest are low. Enrichment and GH scores were used to compare construction protocol for pharmacophore models of varying purposes— such as function specific versus nonspecific ligand identification. Notably, pharmacophore models constructed from ligands of mixed functions (agonists and antagonists) were capable of enriching hitlists with active compounds, and therefore can be used when available sets of known ligands are limited in number. The work in chapter IV applies the protocols developed in Chapters II and III to three target GPCR with important physiological roles—Bombesin receptor subtype 3, GPR88, and GPR52. The pharmacological neighbors of each receptor were examined to identify and prioritize candidate ligands for experimental validation. Additionally, the purchasable and druglike ZINC15 ligand database subset was also examined for hit ligands with the developed pharmacophore models.


Data is provided by the student

Library Comment

Dissertation or thesis originally submitted to ProQuest.


Open Access